Paper Plan

wanshuiyin/auto-claude-code-research-in-sleep

Takes your experimental results, auto-review conclusions, and narrative reports and generates a structured paper outline ready for academic submission. You call it when you say "write outline" or "论文规划" and it consults your figures directory, experiment logs, and review stage files to build a section-by-section plan. The style-ref flag lets you point it at a reference paper (arXiv ID, local PDF, or TeX source) to match that paper's structural rhythm without copying any prose. It emits a gap report showing which sections you have evidence for and which ones are missing data, so you know where placeholders will land when you run paper-write later. Built for ML conference formats like ICLR and NeurIPS but handles IEEE journals and ACM venues too.

skills.sh installs: 558
GitHub stars: 18.9K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add wanshuiyin/auto-claude-code-research-in-sleep --skill paper-plan --agent claude-code

Installs into .claude/skills of the current project.

Files

Paper Plan: From Review Conclusions to Paper Outline

Before each reviewer slot, follow Selectable Reviewers: resolve invocation → .aris/reviewer.json → Claude Code default (Codex MCP) with reviewer_route.py through the canonical helper resolver, preserving this slot's tier, fresh contexts, source/image inputs and deterministic gates. Codex examples/pins below apply only to backend=codex; other routes use their own first-call/reply contract. Save the route and actual returned identity; unavailable selections are REVIEW_UNAVAILABLE. Classify acceptance from actual families/completion/scope. Forward invocation reviewer/model/effort overrides to reviewer-bearing children. Writer selection is separate; follow the writing-worker guidance below when drafting or revising prose.

Generate a structured, section-by-section paper outline from: $ARGUMENTS